I may want to do a test run at the county and place level, for a single year, say 2021, for commuting data in tables DP03, B08006, and C08006. DP03 has data on workers by 6 means of transportation; B08006 has data on workers by 13 means of transportation; C08006 has data on workers by 8 (oops, 9) means of transportation. I think the fewer categories, the less suppression? 

OK. I completed this task: single year 2021 ACS data, US place level, comparing table DP03, B08006, and C08006.

B08006 = 197 US places with non-zero data on workers by means of transportation
C08006 = 427 US places with non-zero data on workers by means of transportation
DP03 = 597 US places with non-zero data on workers by means of transportation
DP03 = 34 US places with values for total population, but N/A for data on workers by means of transportation (Pearland, The Villages, Meridian, etc.)

So, yes, the fewer the categories, the less the suppression. 

Means of Transportation categories for these three tables:

B08006 (13): drive alone; carpool 2; carpool 3; carpool 4+; bus; subway/elevated; commuter rail; light rail; ferry; bicycle; walked; taxi+motorcycle+other; worked at home.

C08006 (9): drive alone; carpool 2; carpool 3; carpool 4+; public transport; bicycle; walked; taxi+motorcycle+other; worked at home.

DP03 (6): drive alone; carpooled; public transport; walked; other; worked at home

Fortuitously, the results for all three tables are consistent (that is, identical). 

Recommendations: Use DP03 if you’re needing data for generalized modes, like carpooling, transit, worked at home.

Use C08006 if you need data on detailed carpooling levels, or bicycle commuters.

Use B08006 if you need data on detailed carpooling, detailed transit submodes; etc.

Use PUMS if you need data separated for taxi or motorcycle or other.

Use Table B08015 if you need an estimate of vehicle driver commuters. (drive alone drivers + carpool drivers). 

That’s all for now.

Chuck



On Sep 16, 2022, at 4:20 PM, Charles Purvis <clpurvis@att.net> wrote:

Hey Ed:

My guess with Redmond, WA and Dublin, CA is that they both popped the magical 65,000 total population barrier in 2021.

Of course, my detailed spreadsheets found enough weirdness to question everything.

Pearland, Texas. Population = 131,448 (2019) and 120,694 (2021). Table DP03 shows workers by mean of commute for 2019 but not for 2021. My R script results match what I can find on data.census.gov. Why the population decline? I’m suspicious.

Meridian, Idaho. Population = 114,161 (2019) and 125,959 (2021). Again, table DP03 shows workers by means of commute for 2019 but not for 2021. Again, matches data.census.gov.

The Villages, Florida. Population 85,377 (2019) and 80,691 (2021). No data on workers by means of transportation to work for either year. This makes sense (?) since The Villages is the largest age 55+ community in the USA. VERY few commuters to be expected. But what happened with total population? A decline? 

Question: Is the 2021 ACS taking into account data on total population, population by age/sex from the now available Census 2020?? I don’t know.

Both the US and States pull both 1 and 52 (states + DC + PR) in my R script for both 2019 and 2021. That’s a relief.

County = 840 in 2019; 841 in 2021… The joined dataset is 852 counties.  A little more messy.

Place = 634 places in 2019; 634 places in 2021; but the joined dataset is 650 places. Some places pop-in; some places are popping-out. Good grief.

PUMA = 2,364 in 2019; 2,364 in 2021; and joined together, still, 2,364. We get the most number of geographic areas in the single-year ACS using PUMAs. And it’s wall-to-wall, shore-to-shining-shore coverage. This is really good to know and to share.

I think we have 2,487 PUMAs based on Census 2020, but I need some verification/ backup from Census Bureau or State Data Centers to check over my analyses.

I may want to do a test run at the county and place level, for a single year, say 2021, for commuting data in tables DP03, B08006, and C08006. DP03 has data on workers by 6 means of transportation; B08006 has data on workers by 13 means of transportation; C08006 has data on workers by 8 means of transportation. I think the fewer categories, the less suppression? 

I went to the White Sox / Athletics game this past Sunday. Dave Stewart’s number retired. Rickey, Dennis, Carney, McGwire, Reggie, Wally Haas, and LaRussa were all there. A’s win the game, too. Fun day in Oakland.

Chuck



On Sep 15, 2022, at 5:25 PM, Ed Christopher <edc@berwyned.com> wrote:

Thanks Chuck. Its always interesting to see the different summaries that people are putting together. Being a small area guy I am sort of wondering what the suppression rule is that is "NAing" the Bethesda and Dublin data in 2019. 

On 9/15/2022 4:31 PM, Charles Purvis wrote:
I’m assembling some of my tweets from today’s efforts. If you’re on twitter, follow my at @charleypurvis

New #ACSdata on workers working at home. Top ten states + US. Using #tidycensus  . What REALLY surprised me is that the US work-at-home share increased from 5.72% in 2019 to 15.82% in 2020 (experimental weights) and FURTHER INCREASED to 17.86 in 2021! Wow. Use Table DP03 for data

<table1_athome.png>

Table 2. Ranking of US Counties #ACSData . DC and neighbor counties; San Francisco; Seattle; NYC; and Atlanta. #tidycensus . These increases are staggering / newsworthy. Had to verify using data.census.gov to be sure!

<table2_athome.png>


Table 3. Work at home by Place (City) of Residence. DC, San Francisco and Seattle suburbs. Redmond and Dublin are super-fast growing burbs. #ACSData #tidycensus . Data matches @kyle_e_walker tweet from this morning. Lots of stories to tell.

<table3_athome.png>

Now focusing on Working-at-Home in the nine-county San Francisco Bay Area. #ACSData #tidycensus . Work at home share increased from 6.5% (2019) to 32.8% (2021) in Bay Area. Wow. A low of 12.8% in Solano to a high of 45.6% in San Francisco County. @MTCBATA

<table4_athome.png>

Number of workers working at home almost quintupled (five-fold increase) in the Bay Area, 2019 to 2021. #ACSData #tidycensus . From doubling plus in Sonoma County to a staggering septupling (seven-fold increase) in Santa Clara County (Silicon Valley) @MTCBATA Pretty wow.

<Table5_athome.png>

The tables are just screenshots of excel tables that I prepared this morning/early afternoon.

That’s all for now.

Chuck






_______________________________________________
CTPP mailing list -- ctpp@listserv.transportation.org
To unsubscribe send an email to ctpp-leave@listserv.transportation.org


-- 
Ed Christopher
708-269-5237
_______________________________________________
CTPP mailing list -- ctpp@listserv.transportation.org
To unsubscribe send an email to ctpp-leave@listserv.transportation.org

_______________________________________________
CTPP mailing list -- ctpp@listserv.transportation.org
To unsubscribe send an email to ctpp-leave@listserv.transportation.org